How AI contributes to improving human interactions. The story behind deeptech Ender Turing

When it comes to business matters, all companies, without exception, want to communicate with customers as smoothly as possible. Scale-up Champions program’s alumni Ender Turing, a deeptech startup founded by Ukrainians, helps them with that. The startup develops an AI-based call center performance management software and speech recognition solution to augment sales, speed up quality control, and skyrocket customer satisfaction.

This spring, Ender Turing raised €200,000 from Estonian Business Angels Network (EstBAN) with ambitious plans to expand to the US market. AIN.Capital publishes the story behind the startup.

Finding a spot for improvement and even coaching

“We started as a technology company, not a problem-solving one. First of all, we created the tech that can process voice communication and create text and analytics based on that text and voice. Three founders were originally from Ukraine, and that’s why in the beginning, we worked with the Ukrainian language,” remembers Olena Iosifova, the CEO of Ender Turing.

The novelty was needed as Google, Amazon, IBM, and others really do not produce great quality speech transcription for call center environments.

Also, there is a catch in communication tools. In the modern hybrid working presence, we do enjoy quite good quality microphones, but when it comes to the call centers, telephones do not yet match the sufficient quality level. Phone connection and phone recording have different audio characteristics, thus making it harder for humans to communicate.

In 2020, Elena Iosifova assembled a team of IT specialists and linguists and created a module that was ahead of all major players in the Ukrainian language.

“It was the moment when we understood the same solution may be applied to other languages. This technology defines us as unique, and also we found a way to work with different dialects,”

the founder notes.

The other companies that work with language analysis algorithms actually work with American English. It is quite problematic for the computer to understand the different dialects. “We can adapt a new language from scratch in two months and if we need to take into account a new dialect, it enhances the quality of recognition to regional accent, we can do it within a month,” Olena says. Today, the technology can analyze 24 European languages.

Toward a deeptech innovation

The success and recognition of the technology developed by the team located in Estonia made them opt for more novelties. As the founder explains, the team made a choice to create a product for business value. This idea blossomed into a creation of a platform for call center performance management, and the primary language recognition technology has become the main part of it.

“Our solution is not about robot communication. We analyze human-to-human interactions and provide a possibility for customer specialists to access their performance on the platform. Thus, people working in the call centers self-learn and self-coach on the best examples of conversations,”

the founder of Ender Turing explains.

AI is a truly precise listener. That is why it understands and exposes our areas for improvement. For example, the startup’s system analyses the communication in an insurance company as the customer service specialist talks to a real customer. The platform reviews the conversation, showing relevant metrics on what was done right and what is not underperformed.

“If the system knows that the person struggles at the start of the conversation, representing himself or the organization, he gets a notification that has to learn to perform better. On the platform, customer support specialists are able to read the best openings of the conversations from those colleagues who do it really well. Sometimes companies measure 8-9 different parts of the conversation, we can track where the performance is dropping. Our clients get the total transparency of the communication on the level of content of communication, not on how many calls were performed or how long those calls were,” pictures the CEO.

Scaling journey from the EU to the States

This spring, during the Latitude59 conference in Tallinn, the €450,000 investment was given out by EstBAN to two companies. Ender Turing was one of them, receiving an investment of €200 000. In parallel, the startup got accepted to one of the top 4 US accelerators: MassChallenge U.S. Early-Stage program. 

“We are now working on the strategy to expand in the US. The raised funds will go toward sales and marketing, penetrating the US market. As we are working with the Spanish market, it is a competitive advantage to reach the US. We can offer the Spanish language analytics, and that is what our main competitors lack,” the founder explains.

In these turbulent times, deeptech sectors where most of these types of novelties are born are facing their challenges. “As the R&D part of our startup is internal, time and money are the most challenging aspects. We did not get a lot of money compared to the other competitors, for instance, the US. For me, as a CEO, it comes as a question of priorities — how to allocate the resources we have, distribute people, money, and attention,” says Olena.

However, she adds that currently, the biggest challenge is the war in Ukraine.

“One of the founders is living in Ukraine, and he cannot leave the country. We also have two colleagues in the country. As relocation is not possible, we wait for Ukraine to win.”